All posts by: Sarah Hansen, M.S. '15

Quantum computers have the potential to revolutionize communications, cybersecurity, and more, by dramatically speeding computation, researchers say. But as Sebastian Deffner notes, “Even quantum computing has shortcomings.”

There may be ways to work around some of quantum computing’s limits, however, further enhancing its speed and other aspects of performance. Deffner, assistant professor of physics at UMBC, and Nathan Myers, a Ph.D. student in Deffner’s research group, will explore techniques to do that with a new three-year, $300,000 grant from the National Science Foundation. And in the process, they just might redefine the fundamental laws of physics.

From paper to the real world

Typically, the quantum systems Deffner’s group (and anyone else) have studied are linear, meaning they are defined by mathematical equations that appear as a line when graphed. However, Deffner says, based on the math, “non-linear systems have very unique capabilities that allow you to circumvent many of the standard problems of linear quantum computing.”

A two-line proof in the grant proposal shows that in theory, non-linear systems can operate much faster than linear quantum systems, and possibly perform better in other ways, too. “Now the question is,” Deffner says, “is that just something you can write on paper, or does it actually play a role in applications?”

Because it doesn’t matter how fast a system could be in theory, if, to go that fast, most of the energy input is released as heat instead of being used for computations. In that case, the total energy required to get anything useful accomplished makes operating such a device impractical in the real world. 

That’s why Deffner’s research group focuses on a burgeoning new field known as quantum thermodynamics—the study of the relationships between heat and other forms of energy in quantum systems. Previously, they developed and refined the idea of the “quantum speed limit,” which quantifies the limits of linear quantum systems. Now they’ll expand that work to non-linear systems.

The beauty of math

So, why is it that most of the quantum systems researchers have studied are linear? And where do non-linear systems come in? 

Well, quantum systems are composed of many particles that are constantly interacting with each other in a linear fashion. Unfortunately, it’s impossible to precisely define and measure all of those interactions using today’s technology. It is possible, however, to approximate the sum of all of those linear interactions. That approximation can be added to the linear equation describing the overall system as a single term. And here’s the catch: That term is non-linear.

By accounting for these particle interactions, “we can potentially get a non-linear speed-up from an underlying linear system. That’s the goal,” Myers says. But—and here’s where the thermodynamics comes in—“Say we get this non-linear system,” Myers says. “What energy cost are we paying for it? Is it worth it?”

So a goal of the new project is to better understand the thermodynamics of these non-linear systems. The findings would help quantify how much faster they might be than linear quantum systems, and if it would be feasible to create and run them in the real world.

A monumental leap

The shift from linear to non-linear systems might seem incremental, but it’s actually monumental. Sadi Carnot, a French engineer, developed the field of thermodynamics in the early 19th century to describe steam engines. Then, in the late 20th century, researchers overhauled his foundational principles to apply them to linear quantum systems.

“We have all these statements of thermodynamics that we just recently formulated for quantum systems,” Deffner says. But non-linear systems are so different that another retrofit won’t be sufficient. “What we need to do is go back to the beginnings of quantum thermodynamics and just redo everything.”

They’ll start by developing three foundational mathematical statements describing the non-linear systems. “Those statements will build the foundation upon which we then can build the whole theory,” Deffner says. They’ll test their work on two major algorithms used in database searches and encryption technology.

“Hopefully our work will lead to a broader exploration of how non-linear systems in general can be used to speed up a whole range of quantum devices, or enhance their performance in other ways,” Myers says.

A global UMBC team

During the pandemic, the research group is on four continents and in “I don’t know how many time zones,” according to Deffner. “I told my students from the beginning that we would have to adapt,” he adds. The lab group stays in touch via Slack, WhatsApp, and weekly group meetings. But the work itself is surprisingly low-tech.

“The majority of it is pen and paper,” Myers says, plus a lot of reading. “Doing good theory work requires having a really broad knowledge base of what’s being done in different disciplines of physics. Real progress happens when someone realizes that a technique or tool from another area is also applicable to their own problem.”

Myers loves the work, and says his entire thesis grew out of a single question he asked Deffner after class one day. Deffner’s response? “That’s a great question. Want to do a project?”

“It all started with one root question that’s then grown a lot of different branches,” he says. “That’s one of the things that makes it so exciting—you’re following this train of thoughts. You’re pulling at the thread in the sweater, it’s unraveling more and more, and you’re seeing how many different things can grow out of this single question. It’s exciting and it’s fun.”

Disruptive discoveries

In the process of having fun, Myers and Deffner will reformulate thermodynamics for non-linear quantum systems, figure out whether building them is feasible, and get an idea of what their perks might be. Deffner’s team just might also come up with something revolutionary. 

Myers tells the story of Carnot, who, in an attempt to optimize the disruptive technology of his day, ended up conceiving the laws of thermodynamics. “By trying to determine the most efficient steam engine possible, he ended up deriving perhaps the most fundamental law in all of physics,” Myers says. 

“Now we’re in a similar position to Carnot,” he suggests. “We have this new disruptive technology that’s emerging—it was steam engines for him, quantum devices for us—so let’s do the same thing, and hope something really incredible falls out of it.”

New 2D materials have the potential to transform technologies, with applications from solar cells to smartphones and wearable electronics, explains UMBC’s Can Ataca, assistant professor of physics. These materials consist of a single layer of atoms bound together in a crystal structure. In fact, they’re so thin that a stack of 10 million of them would only be 1 millimeter thick. And sometimes, Ataca says, less is more. Some 2D materials are more effective and efficient than similar materials that are much thicker.

Despite their advantages, however, 2D materials are currently difficult and expensive to make. That means the scientists trying to create them need to make careful choices about how they invest their time, energy, and funds in development.

New research by Daniel Wines, Ph.D. candidate in physics, and Ataca gives those scientists the information they need to pursue high-impact research in this field. Their theoretical work provides reliable information about which new materials might have desirable properties for a range of applications and could exist in a stable form in nature. In a recent paper published in ACS Applied Materials and Interfaces, they used cutting-edge computer modeling techniques to predict the properties of 2D materials that haven’t yet been made in real life.

“We usually are trying to stay five or so years ahead of experimentalists,” says Wines. That way, they can avoid going down expensive dead ends. “That’s time, effort, and money that they can focus on other things.”

The perfect mix

The new paper focuses on the stability and properties of 2D materials called group III nitrides. These are mixtures of nitrogen and an element from group III on the periodic table, which includes aluminum, gallium, indium, and boron. 

Scientists have already made some of these 2D materials in small quantities. Instead of looking at mixtures of one of the group III elements with nitrogen, however, Wines and Ataca modeled alloys—mixtures including nitrogen and two different group III elements. For example, they predicted the properties of materials made of mostly aluminum, but with some gallium added, or mostly gallium, but with some indium added.

These “in-between” materials might have intermediate properties that could be useful in certain applications. “By doing this alloying, we can say, I have orange light, but I have materials that can absorb red light and yellow light,” Ataca says. “So how can I mix that so that it can absorb the orange light?” Tuning the light absorption capabilities of these materials could improve the efficiency of solar energy systems, for example.

Alloys of the future

Ataca and Wines also looked at the electric and thermoelectric properties of materials. A material has thermoelectric capability if it can generate electricity when one side is cold and the other is hot. The basic group III nitrides have thermoelectric properties, “but at certain concentrations, the thermoelectric properties of alloys are better than the basic group III nitrides,” Ataca says. 

Wines adds, “That’s the main motivation of doing the alloying—the tunability of the properties.”

They also showed that not all of the alloys would be stable in real life. For example, mixtures of aluminum and boron at any concentrations were not stable. However, five different ratios of gallium-aluminum mixtures were stable. 

Once production of the basic group III nitrides becomes more reliable and is scaled up, Wines and Ataca expect scientists to work on engineering the materials for specific applications using their results as a guide.

Back to basics…with supercomputers

Wines and Ataca modeled the materials’ properties using supercomputers. Rather than using experimental data as input for their models, “We are using the basics of quantum mechanics to create these properties. So the good part is we don’t have any experimental biases,” Ataca says. “We’re working on stuff that doesn’t have any experimental evidence before. So this is a trustable approach.”

To get the most accurate results requires huge amounts of computing power and takes a long time. Running their models at the highest accuracy level can take several days. 

“It’s kind of like telling a story,” Wines says. “We go through the most basic level to screen the materials,” which only takes about an hour. “And then we go to the highest levels of accuracy, using the most powerful computers, to find the most accurate parameters possible.”

“I think the beautiful part of these studies is that we started at the basics and we literally went up to the most accurate level in our field,” Ataca adds. “But we can always ask for more.”

A new frontier

They have continued to move forward into uncharted scientific territory. In a different paper, published within a week of the first in ACS Applied Materials and Interfaces, Theodosia Gougousi, professor of physics; Jaron Kropp, Ph.D. ’20, physics; and Ataca demonstrated a way to integrate 2D materials into real devices.

2D materials often need to attach to an electronic circuit within a device. An in-between layer is required to make that connection—and the team found one that works. “We have a molecule that can do this, that can make a connection to the material, in order to use it for external circuit applications,” Ataca says.

This result is a big deal for the implementation of 2D materials. “This work combines fundamental experimental research on the processes that occur on the surface of 2D atomic crystals with detailed computational evaluation of the system,” Gougousi says. “It provides guidance to the device community so they can successfully integrate novel materials into traditional device architectures.” 

Collaboration across disciplines

The theoretical analyses for this work happened in Ataca’s lab, and the experiments happened in Gougousi’s lab. Kropp worked in both groups.

“The project exemplifies the synergy that is required for science and technology development and advancement,” Gougousi says. “It is also a great example of the opportunities that our graduate students have to work on problems of great technological interest, and to develop a broad knowledge basis and a unique set of technical skills.”

Kropp, who is first author on the second paper, is thrilled to have had this research experience.

“2D semiconductors are exciting because they have the potential for applications in non-traditional electronic devices, like wearable or flexible electronics, since they are so thin,” he says. “I was fortunate to have two excellent advisors, because this allowed me to combine the experimental and theoretical work seamlessly. I hope that the results of this work can help other researchers to develop new devices based on 2D materials.”

Header image: Daniel Wines (far right), Jaron Kropp (second from right), Can Ataca (second from left) and other lab members meet. Photo by Marlayna Demond ’11 for UMBC.

The supermassive black holes at the centers of galaxies are the most massive objects in the universe. They range from about 1 million to upwards of 10 billion times the mass of the Sun. Some of these black holes also blast out gigantic, super-heated jets of plasma at nearly the speed of light. The primary way that the jets discharge this powerful motion energy is by converting it into extremely high-energy gamma rays. However, UMBC physics Ph.D. candidate Adam Leah Harvey says, “How exactly this radiation is created is an open question.”

The jet has to discharge its energy somewhere, and previous work doesn’t agree where. The prime candidates are two regions made of gas and light that encircle black holes, called the broad-line region and the molecular torus.

A black hole’s jet has the potential to convert visible and infrared light in either region to high-energy gamma rays by giving away some of its energy. Harvey’s new NASA-funded research sheds light on this controversy by offering strong evidence that the jets mostly release energy in the molecular torus, and not in the broad-line region. The study was published in October in Nature Communications and co-authored by UMBC physicists Markos Georganopoulos and Eileen Meyer.

Far out

The broad-line region is closer to the center of a black hole, at a distance of about 0.3 light-years. The molecular torus is much farther out—more than  3 light-years. While all of these distances seem huge to a non-astronomer, the new work “tells us that we’re getting energy dissipation far away from the black hole at the relevant scales,” Harvey explains.

“The implications are extremely important for our understanding of jets launched by black holes,” Harvey says. Which region primarily absorbs the jet’s energy offers clues to how the jets initially form, pick up speed, and become column-shaped. For example, “It indicates that the jet is not accelerated enough at smaller scales to start to dissipate energy,” Harvey says.

Other researchers have proposed contradictory ideas about the jets’ structure and behavior. Because of the trusted methods Harvey used in their new work, however, they expect the results to be broadly accepted in the scientific community. “The results basically help to constrain those possibilities—those different models—of jet formation.”

On solid footing

To come to their conclusions, Harvey applied a standard statistical technique called “bootstrapping” to data from 62 observations of black hole jets. “A lot of what came before this paper has been very model-dependent. Other papers have made a lot of very specific assumptions, whereas our method is extremely general,” Harvey explains. “There isn’t much to undermine the analysis. It’s well-understood methods, and just using observational data. So the result should be correct.”

A quantity called the seed factor was central to the analysis. The seed factor indicates where the light waves that the jet converts to gamma rays come from. If the conversion happens at the molecular torus, one seed factor is expected. If it happens at the broad-line region, the seed factor will be different.

Georganopolous, associate professor of physics and one of Harvey’s advisors, originally developed the seed factor concept, but “applying the idea of the seed factor had to wait for someone with a lot of perseverance, and this someone was Adam Leah,” Georganopoulos says.

Harvey calculated the seed factors for all 62 observations. They found that the seed factors fell in a normal distribution aligned almost perfectly around the expected value for the molecular torus. That result strongly suggests that the energy from the jet is discharging into light waves in the molecular torus, and not in the broad-line region.

Tangents and searches

Harvey shares that the support of their mentors, Georganopoulos and Meyer, assistant professor of physics, was instrumental to the project’s success. “I think that without them letting me go off on a lot of tangents and searches of how to do things, this would have never gotten to the level that it’s at,” Harvey says. “Because they allowed me to really dig into it, I was able to pull out a lot more from this project.”

Harvey identifies as an “observational astronomer,” but adds, “I’m really more of a data scientist and a statistician than I am a physicist.” And the statistics has been the most exciting part of this work, they say.

“I just think it’s really cool that I was able to figure out methods to create such a strong study of such a weird system that is so removed from my own personal reality.” Harvey says. “It’s going to be fun to see what people do with it.”

Header image: The remnants of a star torn apart by a black hole form a disk around the black hole’s center, while jets eject from either side. Artist’s rendering courtesy of NASA.